Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. Parallel session FAIRMODE 2nd June 2010 2nd June 2010 URBAN EMISSIONS AND PROJECTIONS Rafael Borge 1 , Julio Lumbreras 1 , David de la Paz 1 , M. Encarnación Rodríguez 1 Panagiota Dilara 2 , and Leonor Tarrason 3 1 Laboratory of Environmental Modelling. Technical University of Madrid (UPM) 2 European Commission, Joint Research Centre, Ispra, Italy 3 Norwegian Institute for Air Research. Kjeller, Norway rborge@etsii.upm.es ; jlumbreras@etsii.upm.es
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. OUTLINE 1. Introduction 2. Methodology 3. Results 4. Conclusions
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 3 1. Introduction FAIRMODE flowchart as agreed on 2nd plenary meeting (Nov. 2009) Benchmarking SG(2) + SG(1) SG (4) Combination of SG (3) monitoring and Emission Protocols and Protocols and modelling (data modelling (data inventories and inventories and Tools for assimilation) scenarios benchmarking of AQ models SG(5) Contribution of Urban natural sources and Source Agglomeration apportionment ������������������������ ���
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 4 SG(3) on urban emissions and projections • Background document on the emission needs at local scale • Needs for guidance on emission compilation at urban level – Consistency with national inventories – Top down vs bottom up approches – Use of GIS tools • Urban emission compilation is a key issue at European level • Both guidance and relevant exchange fora are needed • Next step: ‒ Proposal for a framework for the development of emission inventories at local scale • Links to TFEIP/EIONET, NIAM, GEIA, JRC-EDGAR
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 5 assessment of ambient air quality planning and mitigation strategies Air Quality Modelling assessment of the contribution of natural sources, in the AQD road dust and sea salt short-term forecast for threshold exceedances • Uncertainties for Air Quality Models (AQMs) • • Meteorology Meteorology • Modelling system • Boundary and Initial conditions • Emission input • Uncertainties from emission inputs → emission inventories: • Emission data accuracy Consistent emission estimates across the • Temporal disaggregation scales, inventory • Spatial resolution and emission allocation harmonization. Criteria for local scale EI development • Chemical speciation and mass distribution
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 6 1) Emission data accuracy 2) Temporal disaggregation (Wang (Cho et al., 2009) et al., 2010, Kühlwein et al., 2002) 3) Spatial resolution and emission allocation (Mensink et al., 2008, Cheng et al., 2008, Pisoni et al., 2010)
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 7 2. Methodology • To analyse two approaches for different scale emission inventory compilation for an inland city and surroundings : − National calculation using country statistics and some regional data with spatial disaggregation afterwards − Regional calculation using regional data • Compare AQM results (whole year, 1-h resolution) with monitoring data Relate these differences with • Select and analyse a emission compilation methods for number of the dominant source in the grid cell representative stations where the alternative inventories produce important discrepancies Understand reasons for discrepancies, get in AQM results an idea about emission accuracy, and identify options for multi-scale emission inventory harmonization
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 8 • AQM domain including AQ monitoring stations •Same BC and individual profiles for temporal and chemical speciation. Differences in model performance due to: •Emission data accuracy (total figures and sectoral figures) •Emission allocation (source apportionment at grid cell level)
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 9 3. Results • Emission inventory aggregated comparison (INV1 – INV2) SNAP Difference SO 2 NO X NMVOC CH 4 CO NH 3 PM 2,5 PM 10 Absolute -2622 -212 23 -438 -179 -33 -81 1 Relative -100% -42% 750% -41% -42% - -58% -77% Absolute -1729 16269 -1001 -1154 -15749 -811 -882 2 Relative -37% 302% -59% -68% -78% - -83% -83% Absolute -1565 -9824 -1265 209 -3072 -221 74 3 Relative -27% -46% -57% 34% -41% - -52% 13% Absolute 8 11 -1653 566 12 24 4 Relative 6% 6% -44% - 6% - 6% 6% Absolute -2156 27 0 1 5 Relative - - -58% 0% - - 112% 112% Absolute Absolute -2636 -2636 -78 -78 6 Relative - - -4% - - -81% - - Absolute 2605 26601 2573 2067 8124 77 397 -187 7 Relative 963% 53% 17% 267% 9% 11% 10% -4% Absolute 416 4576 287 7 3533 0 -468 -468 8 Relative 67% 53% 32% 22% 75% -18% -71% -71% Absolute 501 310 -3142 -45044 187 765 5 6 9 Relative 1.8E05% 119% -60% -51% 234% 78% 80% 78% Absolute 6 -110 190 -647 -1209 -3782 565 4317 10 Relative 41% -40% 13% -7% -87% -69% 1139% 1285% Absolute 2 -665 8518 -1134 344 -299 11 Relative 117% -97% 53% -96% 117% -98% - - Absolute -2376 36956 -261 -46105 -7455 -3317 -553 2803 TOT Relative -17% 42% 0% -42% -5% -44% -9% 36%
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 10 • Emission allocation (Gridded total NOx emissions according to the inventories considered) INV2 INV1 NO X emissions (t year -1 ) • Largest discrepancies related to road transport and domestic/commcial/institut. heating • Some differences in industry-related combustion processes and off-road mobile sources • Different spatial allocation patterns
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 11 • Emission allocation: a) Source apportionment at grid cell level: SNAP 02 INV2 INV1 SNAP 02 CAM contribution to NO X emissions at grid cell level (%) • Differences: • statistical basis used for activity rates estimation • population as spatial surrogate (uniform emission distribution across a given municipal urban area vs. CORINE land cover population density)
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 12 • Emission allocation: b) Source apportionment at grid cell level: SNAP 07 INV2 INV1 SNAP 07 contribution to NO X emissions at grid cell level (%) • Differences: • discrepancies regarding driving patterns and road classification • differences in mileage estimation per vehicle (daily average intensities and road length vs. prescribed total mileage values depending on vehicle type) • road maps considered
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 13 • AQM results: a) NO 2 annual mean • AQM results: b) NO 2 99.8 th 1-h percentile
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 14 • AQM results: c) NO 2 Mean Bias (ppb) at station level 30 a 25 CAM INV1 20 NAT INV2 15 10 Mean bias (ppb) 5 0 -5 -10 -15 Monitoring stations: -20 -25 A – traffic A – traffic -30 -30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 23 24 26 28 33 34 35 37 39 49 52 B – urban background Station ID 30 30 C – industrial b c 25 25 20 20 15 15 10 10 Mean bias (ppb) Mean bias (ppb) 5 5 0 0 -5 -5 -10 -10 -15 -15 -20 CAM -20 INV1 CAM INV1 NAT -25 INV2 NAT -25 INV2 -30 -30 22 25 29 30 40 41 43 44 45 46 50 53 55 27 31 36 54 Station ID Station ID
Urban Emissions and Projections. Borge, R., Lumbreras, J., de la Paz, D., Rodriguez, M.E., Dilara, P., and Tarrason, L. 15 • Station A (traffic) Station A – INV1 Station A – INV2 E-21 CAM E-21 NAT SNAP-1 100% % SNAP-2 90% % SNAP-3 80% % SNAP-4 70% % SNAP-5 60% % 50% % SNAP-6 40% % SNAP-7 30% % SNAP-8 20% % SNAP-9 10% % SNAP-10 0% % % SNAP-11 CO NOX VOC NH3 SO2 PM10 PM2_5 CO NOX VOC NH3 SO2 PM10 PM2_5 2418 t/y 1093 t/y • INV1 more than double NOx emissions in the corresponding grid cell • SNAP 07 (road traffic) is the predominant source (consistent with station label) • INV2 considers a significant contribution from other sources • NO 2 underestimated with INV2 and overestimated with INV1 similarly • Absolute mean errors (ME) and the correlation coefficient are similar � SNAP 07 emissions largely overestimated in INV1 (excessive contribution of heavy duty vehicles in highway driving patter), although activity ratios are more specific. � Inaccurate secondary EF
Recommend
More recommend
